ex.py

import pandas as pd
import numpy as np
from sklearn.linear_model import LinearRegression
from sklearn.metrics import mean_squared_error
from sklearn.model_selection import train_test_split


def create_lookup_tables(values):
    """
    Create lookup tables for vocabulary
    :param values: List of values to be converted to a dictionary
    :return: A tuple of dicts (vocab_to_int, int_to_vocab)
    """
    vals = sorted(list(set(values)))

    ft_to_int = {word: idx for idx, word in enumerate(vals)}
    int_to_ft = {idx: word for idx, word in enumerate(vals)}
    # return tuple
    return ft_to_int, int_to_ft



if __name__ == '__main__':
    data = pd.read_csv('data - data.csv')
    area_to_int, int_to_area = create_lookup_tables(data['area'].to_list())
    region_to_int, int_to_region = create_lookup_tables(data['region'].to_list())
    print(area_to_int)

    main_data = np.empty((210, 5))
    main_target = np.empty(210)
    # strip out individual values in a row
    for idx, row in data.iterrows():
        area_size = row['area_size']
        num_rooms = row['num_rooms']
        num_bathrooms = row['num_bathrooms']
        target = row['priceGHS']
        area = area_to_int[row['area']]
        region = region_to_int[row['region']]
        entry = np.array([area_size, num_rooms, num_bathrooms, area, region])

        main_data[idx] = entry
        main_target[idx] = target

    X_train, X_test, y_train, y_test = train_test_split(main_data, main_target, random_state=0)

    rfr = LinearRegression()
    rfr.fit(X_train, y_train)
    print(rfr.score(X_train, y_train))
    #prediction on dataset
    x = np.array([304593, 5, 2, 0, 0])
   
    y_pred = rfr.predict(x.reshape(1, -1))
    print(y_pred)